This product is a zip file that contains files that consist of PowerPoint slides synchronized with the audio-recording of the speaker, PDF files of the slides, and audio only (mp3 format) as noted.
An outlier is an observation which deviates from other observations and arouses suspicion. Monitored data are inherently prone to errors caused by instruments malfunction, power surges, disturbance and noise especially with wireless sensors, among other causes. Monitored building energy use data contain anomalous values and gaps of missing readings. The session covers the outliers detection in data-driven modeling, discussing approaches and techniques, including statistical inference, data mining, and artificial intelligence methods, along with an explanation of types of outliers. Cases studies are presented and cover baseline modeling of building energy use, thermal comfort models and whole building energy data screening.
1. Overview of Outlier Detection Techniques with Applications to HVAC&R
T. Agami Reddy, Ph.D., Fellow ASHRAE, Arizona State University, Tempe, AZ
2. A Case Study on the Outliers Detection and Rejection in Data-Driven Baseline Modeling of Building Energy Performance
Bass Abushakra, Ph.D., Member, United States Military Academy, West Point, NY
3. Whole Building Energy Data Quality Assurance through an Energy Balance Loads Approach
Juan-Carlos Baltazar, Ph.D., BEMP, Member, Texas A&M University, College Station, TX
4. Machine Learning for Anomaly Detection in Subjective Thermal Comfort Votes
Zhe Wang, Ph.D., Lawrence Berkeley National Laboratory, Berkeley, CA
| File Size : | 1
file
, 96 MB |
| Note : | This product is unavailable in Russia, Belarus |
| Product Code(s) : | D-OR20Sem61 |
| Published : | 2020 |
